In [3]:
# Get the class labels and their corresponding indices
class_labels = train_generator.class_indices
# Create a dictionary to store the first image path from each class
class_image_paths = {}
# Iterate over the train generator to find the first image path from each class
for i in range(len(train_generator)):
# Get the image and label from the generator
image, label = train_generator[i]
# Get the class label index
class_index = label.argmax()
# Get the class label name
class_name = list(class_labels.keys())[class_index]
# If the class is not yet added to the dictionary, store its image path
if class_name not in class_image_paths:
image_path = train_generator.filepaths[i]
class_image_paths[class_name] = image_path
# If images from all classes are found, break the loop
if len(class_image_paths) == len(class_labels):
break
# Create a grid of subplots
fig, axs = plt.subplots(len(class_image_paths) // 4 + 1, 4, figsize=(15, 15))
# Flatten the axs array for easier indexing
axs = axs.flatten()
# Display the images from each class
for i, (class_name, image_path) in enumerate(class_image_paths.items()):
# Load the image
image = plt.imread(image_path)
# Display the image in the corresponding subplot
axs[i].imshow(image)
axs[i].set_title(class_name)
axs[i].axis('off')
# Break the loop if all subplots are filled
if i + 1 == len(axs):
break
# Hide any remaining empty subplots
for j in range(len(class_image_paths), len(axs)):
axs[j].axis('off')
# Adjust the spacing between subplots
plt.tight_layout()
# Show the plots
plt.show()
